摘要
Inthispaper,anewbiasestimationmethodisproposedandappliedinaregionalensembleKalmanfilter(EnKF)basedontheWeatherResearchandForecasting(WRF)Model.Themethodisbasedonahomogeneouslinearbiasmodel,andthemodelbiasisestimatedusingstatisticsateachassimilationcycle,whichisdifferentfromthestateaugmentationmethodsproposedinpreviousliteratures.Thenewmethodprovidesagoodestimationforthemodelbiasofsomespecificvariables,suchassealevelpressure(SLP).AseriesofnumericalexperimentswithEnKFareperformedtoexaminethenewmethodunderasevereweathercondition.Resultsshowthepositiveeffectofthemethodontheforecastingofcirculationpatternandmeso-scalesystems,andthereductionofanalysiserrors.ThebackgrounderrorcovariancestructuresofsurfacevariablesandtheeffectsofmodelsystembiasonEnKFarealsostudiedundertheerrorcovariancestructuresandanewconcept‘correlationscale’isintroduced.However,thenewmethodneedsfurtherevaluationwithmorecasesofassimilation.
出版日期
2013年03月13日（中国期刊网平台首次上网日期，不代表论文的发表时间）
